Activity Coefficients in Aqueous Zinc Chloride-Hydrochloric Acid

Activity Coefficients in Aqueous Zinc Chloride-Hydrochloric Acid Solutions, and their Application to Cation Exchange Data1. Norman J. Meyer, William J...
0 downloads 0 Views 633KB Size
N. J. MEYER,W. J. ARGERSINGER, JR.,

1024

in discussing kinetic effects of added "inert" electrolytes conclude that from kinetic data only the choice is difficult, if not impossible. The very effective catalysis by Co(NH&+++ is analogous to the catalysis of the bromoacetate-thiosulfate reaction?6by La+++ and is reasonable by either interpretation. We believe the catalysis by Fe(CN)6--- involves the rapid reversible reaction?'

+ Fe( CN)e----

ki k?

+ Fe(CN)B---

A. W. DAVIDSON

K'

=i

-200

(5)

(26) V. K . La Mer a n d M. E. Kamner, THISJOURNAL,57, 2662 (1935). (27) T h e value of t h e equilibrium constant was estimated from t h e formal potentialz8 of the Fe(CN)e-----Fe(Cr\')s--' couple in 0.083 f NarSOc a n d t h e E o valueaQ of t h e MnO,---MnOccouple, both a t N o correction for temperature was made. 2.5'. (28) I. M. Kolthoff a n d W. J Tomsicek, J . Phys. Chem., 39, 8-45 (1835). (29) W. hi. Latimer, "Oxidation Potentials," Prentice-Hall, Inc., S e w York, N. Y . , 1932, p. 239.

[CONTRIBUTION FROM THE CHEXICAL

VoI. T9

The rate under the experimental conditions is given by R = 11S0(~vIn04--)(hInOI-) = 710(MnOd--)(MnOa-)

$-

kq(MnOl--)( Fe( cx)O---)

kl = K'k, = (-200)(47)

Mn04-

MnOa--

AND

=

-1041M-*sec.-]

This value of K1 is approximately the value we estimated from visual observations of the color changes that occur when Mn04- and Fe(CN)c---are mixed in 0.lfKOH a t 5". Acknowledgment.-We wish t o thank Professor E. L. King of the University of Wisconsin for reading the manuscript and making many valuable comments. ST. LOUIS,MISSOURI

LABORATORY O F THE

UNIVERSITY O F KAXSAS]

Activity Coefficients in Aqueous Zinc Chloride-Hydrochloric Acid Solutions, and their Application to Cation Exchange Datal BY NORMAN J. MEYER,WILLIAMJ. ARGERSINGER, JR., AND ARTHURW. DAVIDSON RECEIVED OCTOBER 4, 1936 The equilibrium quotient for the zinc-hydrogen exchange on Dowex 50 in aqueous chloride solutions has been determined as a function of resin composition a t constant total solution ionic strengths of 0.5 and 1.0 m. Approximate values of the solution activity coefficient ratios, obtained from electromotive force measurements on mixed hydrochloric acid-zinc chloride solutions and the application of Harned's rule, are used to compute the thermodynamic equilibrium constant for the exchange process and the activity coefficients of the resin components. The values obtained for the equilibrium constant are 10.2 at 0.5 m ionic strength and 11.7 a t 1.0 m ionic strength.

Introduction Equilibrium in the process of cation exchange on a synthetic resin such as Dowex 50 may be formulated in terms of a thermodynamic equilibrium constant which includes the activity coefficients of the components of the resin phase. The method of calculation of these quantities which has been developed in several previous papers from this Lab ~ r a t o r y ? -has ~ been applied t o a number of exchange systems involving two univalent cations?-9 and to a smaller number of exchange systems involving either two bivalent ions or one bivalent ion and one univalent ion.I0 At least one earlier (1) From p a r t of a thesis submitted b y Norman J . Meyer in partial fulfillment of t h e requirements for t h e degree of Doctor of Philosophy, University of Kansas, 1956. (?) W. J. Argersinger, Jr., A. W. Davidson a n d 0. D. Bonner, Trans. Kans. A c a d . Sci.. 53, 404 (1950). (3) W. 1. Argersinger, Jr., a n d A. W. Davidson, J . P h y s . Chem., 66, 90 (1952). (4) 0. D. Bonner, W. J. Argersinger, Jr., and A. W. Davidson, THIS J O U K N A L , 74, 1044 (1952). ( 5 ) A. W. Davidson and W. J. Argersinger, Jr., Ann. N . Y. Acad. Sci., 57, 105 (19531. ( 6 ) 0. D. Bonner and V. R h e t t , J . P h y s . Chcm., 57, 254 (1953). (7) 0. D . Bonner, i b i d . , 59, 719 (1955). (8) G. E. Wilson, A. W. Davidson and W. J. Argersinger, Jr., THIS I O U R N A L , 76, 3824 (1954). (9) E. W. Baumann and W. J. Argersinger, Jr., ibid., 78, 1130 (1956). (10) 0. D . Bonner a n d F. L. Livingston, J. P h y s . Chcm., 60, 530 (1956).

study" was made of systems involving the exchange of bivalent ions on Dowex 50, but in this instance approximate methods were used for the computation of the exchange equilibrium constant, and resin activity coefficients were ignored. The computation of the exchange equilibrium constant requires, in addition to the exchange data themselves, a knowledge of the activity coefficients of the electrolyte solutes in the aqueous phase-or, more precisely, of the appropriate ratio of these coefficients. If the aqueous solutions are sufficiently dilute, this ratio may be taken as unity6*'; in somewhat more concentrated solutions, the ratio may be estimated from a modified form of the ionic strength principle.* In general, however, it seems preferable t o use experimentally determined values of the activity coefficient ratios if such are available. Thus in the case of the silver-hydrogen exchange on Dowex 50, the values of the ratios of activity coefficients of silver nitrate and nitric acid were obtained from independent electromotive force measurements.'? The direct determination of the activity coefficients of both electrolyte solutes in a mixed aqueous solution is extremely difficult; it has been ac(11) W. K. Lowen, R. W. Stoenner, W. J. Argersinger, Jr.. A. W. Davidson a n d D. 37.Hume, THIS J O U R N A L , 73, 2666 (1951). (12) 0. D. Bonner, A. W. Davidson and W . J. Argersinger, Jr.. ibid., 74, 1047 (1952).

complished by Harned and CookI3 for hydroxidechloride mixtures by means of electromotive force measurements on suitably contrived cells. Somewhat less directly the coefficients may be determined from vapor pressure data for the mixed solutions, as suggested by McKayI4and by McKay and Perring,I5 and accomplished by Bonner and HollandL6for aqueous solutions of p-toluenesulfonic acid and its sodium salt, and by McCoy and Wallace” for aqueous solutions of potassium chloride and bromide. In a few instances i t is possible to determine directly the activity coefficient ratios from electromotive force measurements on cells containing electrodes reversible with respect to the two different c a t i o n ~ , ~and ~ ~ ~to~ ~ compute 1Q therefrom the activity coefficients of the two solutes.20 It has been shown by Harned and his co-workers that in many aqueous mixtures of hydrochloric acid with a metal chloride at constant total ionic strength, the activity coefficient of the acid is a linear function of its ionic strength.21 It has been inferred that in such cases the activity coefficient of the metal chloride is also a linear function of its ionic strength a t constant total ionic strength, and experimental results by Harned and Garyz2 seem to justify this assumption a t least for higher concentrations in systems containing barium, strontium, aluminum or cerium chloride. Thus for those cation-exchange systems involving hydrogen ion and the ion of a metal more reactive than hydrogen, in chloride solution, the requisite activity coefficient ratios may be obtained from electromotive force measurements designed to yield the activity coefficient of the acid, and the assumption that Harned’s rule holds for both electrolyte solutes. I n this paper data are presented for the exchange of zinc and hydrogen ions on Dowex 50 a t 25’ Zn++

+ 2HR +ZnRz + 2 H +

(1)

for which the equilibrium constant K is defined as

- a2H+ aZn*

aznRI a2HR

1025

ZINC-HYDROGEN EXCHANGE IN CHLORIDE SOLUTIONS

March 5, 1957

- mzH+ WZZn“

ylHCL T’ZnGIs

N z ~ RX, f ~ l XN

2

~f2HR~ (2)

in which a represents activity, m molality and y mean molal activity coefficient in the aqueous solution; N represents mole fraction in the resin and f resin activity coefficient on a mole fraction basis. It is convenient also to define the equilibrium quotient K m and the apparent equilibrium constant K ,

(13) H . S. Harned and M. A. Cook, ibid., 69, 1890 (1937). (14) H. A. C . McKay, Nature, 169, 464 (1952). (15) H. A. C. McKay and J. K. Perring, Trans. Faraday Soc., 49, 163 (1953). (16) 0. D. Bonner and V. F. Holland, THISJOURNAL, 11, 5828 (1955). (17) W. 11. McCoy and W. E. Wallace, i b i d . , I S , 1820 (1956). (18) V. F. Holland and 0 . D. Bonner, i b i d . , 77, 5833 (1955). (19) 0 . D. Bonner and F. A. Unietis, ibid., 1 6 , 5111 (1953). (20) W. J . Argersinger, Jr., J. P h y s . Chem., 68, 792 (1954). (21) H. S. Harned and B . B. Owen, “The Physical Chemistry of Electrolytic Solutions,” Reinhold Publ. Corp., New York, N . Y . , 1950, p. 459. ( 2 2 ) H. S. Harned and R. Gary, THISJ O U R N A L , 76, 5924 (1954); 11, 1995, 4695 (1955).

so that K = Ka-f Z n R i

(4) f 2HR This paper presents also data for the determination of approximate values of YHCl from electromotive force measurements on the cell

H2(1atm.)/HCl ( m = p ) ,ZnC12( m =

1 / ~ p1 (

- x))/AgCl-Ag

in which p is the total ionic strength and x the ionic strength fraction of hydrochloric acid. The electromotive force of such a cell is given by RT x(2 x) E = Eo - F In pZ 3 y12 (5)

+

where y1= YHCl and the other symbols have their usual meanings.

Experimental Methods Ion-exchange Measurements.-The sodium form of Dowex 50 containing 8% DVB was completely converted t o either the hydrogen or the zinc form by treatment with concentrated hydrochloric acid or concentrated zinc chloride solution; the products were thoroughly washed with water and then air-dried. Stock solutions of zinc chloride and of hydrochloric acid were prepared with conductivity water and Bakers Analyzed Reagent grade or Mallinckrodt Analytical Reagent grade chemicals. The acid solution was analyzed by standard methods, and the zinc chloride solution was analyzed for both zinc and chloride, the former both gravimetrically by the sulfate methodza and volumetrically by the ferrocyanide method,24the latter gravimetrically as silver chloride. The stock solutions were mixed in predetermined proportions, diluted with conductivity water, and stored in glass-stoppered bottles. Triplicate samples of pure hydrogen or zinc resin weighing approximately two grams were shaken for one t o two hours with portions of a selected mixed electrolyte solution. The process was twice repeated and then the resin was transferred t o a small exchange column, approximately 1 cm. in diameter, placed in a thermostat maintained a t 25 & 0.1’. After thermal equilibrium had been reached, fresh mixed solution of the same composition as that earlier used was allowed t o percolate through the resin in the column. Although analyses showed only a negligible change in composition of the solution after the passage of about 100 ml. of solution, 200-300 ml. more was passed through t o ensure the attainment of equilibrium. In this manner it was possible t o obtain equilibrium data a t constant total ionic strength and specified solute composition of the equilibrium aqueous solution. The analysis of the equilibrium resin was inevitably complicated by the interference of zinc ion with the accurate determination of hydrogen ion by titration. In practice, the difficulty could be overcome only by the application of empirical corrections determined from the analysis of mixtures of known compositions. The equilibrium resin was rapidly but thoroughly washed with water and then exhaustively exchanged with a concentrated solution of an eluting electrolyte. Attempts were made t o analyze the resin by elution with magnesium sulfate solution followed by analysis of the eluate for both hydrogen and zinc ions, but by this method the requisite accuracy could not be attained for either ion, primarily because of the high concentration of magnesium sulfate. The method finally adopted involved the determination of one ion and of the total number of equivalents of resin, the number of equivalents of the second ion being obtained as the difference. I n a few cases, the equilibrium resin was completely eluted with hydrochloric acid, the eluate was analyzed for zinc by the ferrocyanide method, and the exchanged resin was washed and then eluted with sodium chloride or nitrate solution to yield an eluate which was titrated t o give the total number of equivalents of exchanged ions. Most commonly, however, because the resin so strongly prefers zinc over hydrogen, the equilibrium resin was completely eluted with sodium (23) W. W. Scott, “Standard Methods of Chemical Analysis,” Vol. 1, D. Van Nostrand Co., New York, N. Y.,1939. (24) I. M . Kolthoff and E. B. Sandell, “Textbook of Quantitative Inorganic Analysis,” The Macmillan Co., New York, N. Y . , 1948.

chloride solution and the eluate titrated with standard base to give the number of equivalents of hydrogen; the exchanged resin was converted t o the hydrogen form by treatment with hydrochloric acid, thoroughly washed, and then again eluted with sodium chloride solution. This second eluate was titrated with standard base t o give the total number of equivalents of equilibrium resin. The methods were tested on known samples; only very small corrections to the titrations were required because of the added electrolyte present. Electromotive Force Measurements.-The materials and stock solutions were the same as those used for the ion-exchange investigation. Silver-silver chloride electrodes were prepared as recommended by Rule and Lah'Ierz5 or by Shedlovsky and MacInnes.26 The potentials of these electrodes were compared in 0.1 X hydrochloric acid solution and found to agree within 0.05 mv. The electrodes were stored in 0.1 114 hydrochloric acid. Hydrogen electrodes were prepared as recommended by Popoff, Kunz and Sn0w.l' The various combinations of hydrogen and silver-silver chloride electrodes used in the study of mixed solutions were always checked before such use, and usually afterwards zs well, in dilute hydrochloric acid solutions; they were coiisidered satisfactory if the potentials were sensibly constant and within 0.05 mv. of the appropriate values of Harned and Ehlers2*for the acid solution. The electrodes in pairs and the solutions were placed in an all-glass cell which was immersed in a thermostat a t 25 =k 0.1". The solution was freed of oxygen by means of nitrogen gas previously saturated with water vapor from a soiution of the same composition, and was maintained oxygen-free through the special design of the cell. After thermal equilibrium had been attained, potential measurements were made with a Leeds and Northrup Type K2 potentiometer and a Leeds and The measured poNorthrup No. 2430 galvanometer. tentials were corrected to a hydrogen pressure of one atmosphere by means of the usual (approximate) corrections.*9 At least three independent measurements were made for each solution; the maximum difference between results for a given solution was 0.1 mv., and the average deviation from the mean was about 0.05 mv.

Results and Calculations Activity Coefficients.-Table I contains the values of log y1 calculated by means of eq. 5 froin e.m.f. measurements on mixed solutions at constant total ionic strengths of 0.2, 0.5 and 1.0 m. A t 25" Eo is taken as 0.22236 v. and R T / F as 0.05914 v. The final figure in each tabulated value of log y1 is uncertain but is included for the sake of completeness. TABLE

Aci IVITU ComPIcrtims X

0,087 ,182 ,286 ,400 ,626 ,667 ,824 927 lL

A. W. DAVIDSON

S . J. MEYER,b-. J. ARGERSINGER, JK., A N D

1026

HC1 log Y l

OF

* = 0.2 -0.1242 1238 ,1206 .11w ,1177 ,1166 ,1148 1133 ,1134"

I IS p

IICl-ZiiCl? SoLwIoss 0.5

-0.1484 ,1406 ,1372 ,1372 1285 1273 ,124-i 1210 ,1208"

1.000 \'dues of Harned and Ehlers (ref. 28).

* = 1.0 -0.1569 ,1491 ,1407 ,1373 ,1268

,1147 ,1087 .(I999 ,0920"

log

i l =

10gic0, -

~

) =

e12

0.0495 (0.2)( 1 0.0.547 (0.5) (1 0.0680(1.0) ( I

- X) - Y'I

(1;)

- X,I

Although Harned's rule is known not to hold exactly at low ionic strengths, we shall nevertheless apply it here to the second solute, zinc chloride, in order to obtain approximate values of the activity coefficient for use in ion-exchange calculations. Harned and Gary:!2have shown that if log

?'2

= log

(7)

- e2:p>:

then for the present case 1

021

=

2a,2 -

;j:, -.ap

-

["?I

cjz

- 11

(8)

in which 41 and $2 are the osmotic coeRicients of hydrochloric acid and zinc chloride, respectively, in their own pure solutions a t the same ionic strength as the mixed solution. The values of a2> so computed are shown in Table 11, together with values of the osmotic and activity coefficients of the two solutes obtained by interpolation from the data compiled by Robinson and Stokes,30and also the value of 2a12 CY^^, which should be constant a t all ionic strengths if Harned's rule holds precisely. The several values of 2alz a12 are, it is true, only roughly in agreement: i t is believed, however, that the constancy of this quantity constitutes an extremely sensitive test of the validity of Harned's rule. The degree of divergence here observed would scarcely justify the discarding of this rule as a means of obtaining approxiniate results for y2.

+

+

I1 I